AI & Data Processing Addendum (DPA)
This AI & Data Processing Addendum ("Addendum") forms an integral part of the agreements governing the use of Amdura™ ("we," "our," or "us") digital platforms, agency solutions, cross-cultural UX research software, and artificial intelligence interfaces, including our conversational assistant, Aries™.
Amdura™ is committed to building safe, human-first, and highly secure technology. This document explicitly defines our technical safeguards, zero-training privacy guarantees, and ethical safety standards across all software and AI tools developed by Amdura™.
1. The Zero Model Training Guarantee
Amdura™ respects the strict confidentiality of our clients, website visitors, research participants, and platform users. We enforce a uncompromising policy regarding artificial intelligence model training:
- No Public Model Training: Your prompts, text interactions, uploaded files, branding assets, and personal inputs are never used to train public AI models or foundational datasets.
- Enterprise Endpoint Isolation: All AI features integrated into Amdura™ platforms or client projects (including integrations powered by Google Gemini Pro and OpenAI enterprise APIs) utilize paid commercial endpoints configured strictly with zero-data-retention or zero-training settings.
- Data Ownership: You retain complete, unencumbered ownership of all data submitted to or generated by AI tools built or deployed by Amdura™.
2. AI Safety, Harm Reduction & Ethical Design
Safety is the core pillar of all software engineering at Amdura™. Whether interacting with Aries™ or custom algorithms deployed for enterprise clients, our technology operates under strict ethical guardrails:
- Zero Harm Policy: Aries™ and all Amdura™ software are architected to prevent harm. Our systems strictly prohibit the generation or propagation of abusive content, malicious code, hate speech, or dangerous instructions.
- Dark Pattern Elimination: We do not design or utilize manipulative user interfaces, deceptive AI prompting, or psychological dark patterns intended to coerce user consent or exploit vulnerability.
- Human-Centric Guardrails: Autonomous agents and algorithm logic developed by Amdura™ include system-level system prompts, safety filters, and human oversight mechanisms to ensure reliable, safe execution.
3. Data Processing & Security Standards
When personal data or client content is processed through Amdura™ web services or custom software:
- Encryption Baseline: All data transmitted to or from our servers or AI APIs is protected using Industry-standard Transport Layer Security (TLS 1.3) in transit and Advanced Encryption Standard (AES-256) at rest.
- Subprocessor Transparency: We only pass data to vetted, enterprise-grade cloud vendors (such as Hostinger, Cloudflare, Google Workspace, and Stripe) as detailed in our public Subprocessors List.
- Minimal Data Retention: Intermediary conversational logs and session states are automatically purged or anonymized to prevent unnecessary data persistence.
4. Disclaimers & AI Output Verification
Artificial intelligence language models generate probabilistic outputs. While Amdura™ implements rigorous prompt engineering and system-level validation to optimize accuracy:
- AI-generated text, research summaries, and conversational responses provided by Aries™ or integrated AI plugins are intended as informative tools and should be reviewed by qualified human operators prior to making critical legal, financial, or medical decisions.
- Amdura™ is not liable for errors, omissions, or third-party interpretations resulting directly from automated output generation where reasonable human review was omitted.
5. Contact & Privacy Inquiries
For questions regarding our AI safety protocols, data processing practices, or to request a signed copy of this DPA for your enterprise account, please reach out to our team:
Amdura™ Operations
Lawrence, MA 01840, United States
Email: legal@amdigitaldesigns.com